The behavior of alpha2 plasmin inhibitor in fibrinolytic states

The behavior of alpha2-plasmin inhibitor in

fibrinolytic states.

N Aoki, … , M Matsuda, K Tachiya

J Clin Invest.

1977;

60(2)

:361-369.

https://doi.org/10.1172/JCI108784

.

Human plasma alpha2-plasmin inhibitor in fibrinolytic states was studied using

immunochemical methods and radioiodinated plasminogen. The concentration and activity

of plasma alpha2-plasmin inhibitor decreased when urokinase was added to plasma in vitro

or infused intravenously in man. The decrease was associated with the appearance of

plasmin-alpha2-plasmin inhibitor complex which subsequently disappeared from the

circulation in a short time. A decrease of other major inhibitors, such as

alpha2-macroglobulin and alpha1-antitrypsin, was not observed when the amount of urokinase

added or infused was relatively small, and conversion of plasminogen to plasmin was not

extensive. The formation of plasmin-alpha2-macroglobulin complex was observed only

when plasma plasminogen was activated with a larger amount of urokinase, and after most

of the alpha2-plasmin inhibitor was consumed by forming complexes with plasmin. The

formation of plasmin-alpha1-antitrypsin complex was not observed even in the highly

activated plasma unless exogenous plasmin was added to the plasma. alpha2-Plasmin

inhibitor was the only inhibitor of which the concentration in plasma was significantly

decreased in patients with disseminated intravascular coagulation and fibrinolysis among

the major plasmin inhibitors in plasma. The most reactive inhibitor for regulating plasma

fibrinolysis very likely is alpha2-plasmin inhibitor.

Research Article

Find the latest version:

The

Behavior of a2-Plasmin

Inhibitor

in

Fibrinolytic

States

NOBuo AoKi, MASAAKI

MOROI, MICHIO

MATSUDA, and KAZUKO TACHIYA

From the Department of Medicine and the Institute of Hematology, Jichi Medical School, Minamikawachi-Machi, Tochigi-ken 329-04,Japan

A

B S

T R A C T

Human plasma

a2-plasmin inhibitor in

fibrinolytic states was studied using immunochemical

methods and radioiodinated plasminogen. The

con-centration

and activity of plasma

a2-plasmin

in-hibitor decreased when urokinase was added to plasma

in

vitro or infused intravenously in man. The

de-crease was

associated with the appearance

of

plasmin-a2-plasmin inhibitor complex which

sub-sequently disappeared from the circulation

in

a

short

time. A

decrease of other major inhibitors, such as

a2-macroglobulin and

a1-antitrypsin, was not observed

when the amount of urokinase added or infused was

relatively small, and conversion

of plasminogen to

plasmin

was not extensive.

The formation of

plasmin-a2-macroglobulin complex was

observed only

when

plasma

plasminogen

was

activated with a

larger amount

of

urokinase,

and

after

most

of the a2-plasmin

in-hibitor was consumed by

forming

complexes with

plasmin. The formation

of

plasmin-al-antitrypsin

complex

was not

observed

even in

the

highly

acti-vated

plasma unless exogenous

plasmin

was

added

to the

plasma. a2-Plasmin inhibitor

was

the

only

inhibitor

of which the concentration

in

plasma

was

significantly decreased

in

patients with disseminated

intravascular coagulation and

fibrinolysis among the

major

plasmin inhibitors

in

plasma. The

most reactive

inhibitor

for regulating plasma fibrinolysis very

likely

is

a2-plasmin

inhibitor.

INTRODUCTION

A

plasma

fraction

is

known which inhibits

plasmin-ogen activator-induced clot

lysis

efficiently

but

pos-Part ofthisdata was presentedat the Intemational Society

ofHematology Meeting,8September1976.Theresearch was

carried out according to the provisions of the Declaration of Helsinki, and informed consent was obtained from each

one ofthepatients.

Receivedfor publication23September1976andinrevised form 1April 1977.

sesses

only a small part

of the whole capacity of

plasma to neutralize plasmin activity (1, 2). This

fraction

is

called

antiactivator, since it was

found

to

be

different from the major known plasmin inhibitors

such as

a2-macroglobulin (a2-M)l and a1-antitrypsin

(a,-AT),

both which strongly inhibit plasmin on incubation but

exert

little effect on activator-induced clot lysis (1, 3).

This antiactivator protein was purified, and its

proper-ties were

investigated (4). The antiactivator

was

ulti-mately

found to be primarily a plasmin inhibitor

be-longing to a2-globulin,

and the

term.

a2-plasmin

inhibitor (a2-PI) or a2-proteinase inhibitor is

con-sidered to

be appropriate. a2-PI

was

shown

to

be

dif-ferent

from a2-M,

CI

inactivator,

inter-a-trypsin

in-hibitor,

a1-antichymotrypsin,

a,-AT,

and antithrombin

III

(4). a2-PI

inhibits plasmin

almost

instantaneously,

and the inhibition of

plasminogen

activation

by

the

inhibitor was found to be mainly

due to the very

rapid inactivation of plasmin

formed (4).

In

this paper,

we

consider that this unique

biological property of

a2-PI has

an

important role

in

regulating

fibrinolysis

invivo.

METHODS

Plasma. Human blood was collected from antecubital veins into a 0.1-volof3.8%trisodiumcitrateandwas

centri-fuged at 2,000 g at tip for 20 min to prepare platelet-poorplasma.

Preparation of purifiedplasminogen and plasmin. Plas-minogen was prepared from human freshplasma according

tothe method ofBrockwayand Castellino (5). Plasminogen fraction2, whichwaseluted after fraction 1,wasused for all the studies.Plasmin wasprepared byactivatingplasminogen

with urokinase-coupled Sepharose (Pharmacia Fine Chem-icals, Div. ofPharmacia, Inc., Piscataway, N.J.) according

tothe methodpreviously described(4).Activity of

plasmino-gen and plasmin wasassayed accordingtothe caseinolytic

method of Robbins and Summaria (6). Protein

concentra-'Abbreviations used in this paper:a2-M,

a2-macroglobulin;

a2-PI, a2-plasmin inhibitor; a1-AT,

a,-antitrypsin;

DIC, dis-seminated intravascular coagulation.

tion of purified plasminogen was determined by measuring the absorbance of the plasminogen solution (pH 7.4) at 280 nm and by converting absorbance to protein concentration usingE19m= 17.0 for purified plasminogen (7).

Urokinasepreparations. Urokinase preparations used for intravenous infusion were obtained from Green CrossCorp.,

Osaka and from Mochida Pharmaceutical Co., Tokyo. Ac-tivity was expressed by intemational units based on the standard preparations supplied by the World Health Organ-ization. The unit is nearly equivalent to the Committee of Thrombolytic Agents unit.

Assayfor plasma plasminogenand fibrinogen. Plasmin-ogen in plasma was measured after destruction of

anti-plasmin by acidification (8). All plasmas to be assayed

were first acidified by the addition of the equal volume of 0.166 N HCI, allowed to stand for 15 min, reneutralized with 0.166 N NaOH, and then subjected to the caseinolytic method (6). The results were expressed in Remmert and Cohen casein units (9). The fibrinogen content of plasma wasassayed by the method of Ratnoff and Menzie (10).

Immunochemicalmethods. Rabbit antisera against human

a,-AT

anda2-M were obtained from Behring-Werke AG,

Mar-burg/Lahn, WestGermany. Rabbit antiserum against human a2-PI was prepared by immunizing rabbits with a purified preparation of the inhibitor (4). About1mgof protein in 1 ml

of 0.05 M Tris-HCl, pH 7.4, containing 0.15 M NaCl

was mixed completely with an equal volume of Freund's complete adjuvant and was injected in equally divided volumes intracutaneously into each footpad, shoulder, and

thigh of a rabbit. 3 wk later, the injection was repeated

with 0.5 mg protein. Immune sera were harvested2-5 wk

afterthesecondinjection,andwereabsorbedwith an a2-PI-free plasmafraction instep 4 of the purification procedures

ofthe inhibitor described previously (4). The fraction had

been passed through the plasminogen-coupled Sepharose

column repeatedlytoremove a2-PI completely. 1 vol of this plasma fraction, absorbancy of whichwas 17 at 280nm,was

added to 40 vol of antiserum, incubated at 37°C for 15

min, and stored at 4°C for 48h. The resulting precipitate was removed by centrifugation anddiscarded.Thesupernate

was storedina small aliquot with 0.1% ofsodium azide as a preservative. Antiserum thus obtained was shown to be

monospecific by double immunodiffusion and

immuno-electrophoresis aspreviouslydescribed(4). The IgG fraction

ofantisera wasobtainedbyDEAE-cellulosechromatography

(11). Rabbitantiserum against humanplasminogen was

pre-pared by immunizing rabbits with a purified preparation of plasminogen.Immunizationprocedurewasessentially the

same asthe one usedforantiserum against a2-PI described

above.About10mg of proteinwasused for the first injection and 1.5 mg for the second injection. Immune sera were ab-sorbed with plasminogen-free human plasma which had

passed through a column oflysine-Sepharose (4). 1 vol of

fivefold-diluted plasminogen-freeplasma was added to9vol

ofantiserum, andthe precipitate formedwasremoved. Theconcentration ofa2-PI in human plasma or serum was

determined bythe one-dimensional method of Laurell (12) using1% antiserum inagarose (0.8%inbarbitalbuffer,pH 8.6, ionic strength 0.05). 10

_p1

of antigen solution (purified

inhibitor, plasma, or serum) was applied to the electro-phoresis which wasrunwithaconstant current of1 mA/cm

width for 7h. Dilutions ofknown amounts of the purified

inhibitor were made with Tris-buffered saline (Tris 0.05 M, NaCl 0.15 M, pH 7.4) containing 3% wt/vol bovine serum

albumin and used for calibration. There was a linear

rela-tionship between peak heights and amounts of antigen ap-plied (Fig. 1). The purified inhibitor used for calibration

was the one which had the highest specific activity, 1,933

U/ml per

A280

(4), in a series of preparations and appeared as a single band on sodium dodecyl sulfate polyacrylamide gel electrophoresis. The protein concentrations of the puri-fied inhibitor were determined by measuring the absorbance of the inhibitor solution (pH 7.4) at 280 nm and converting absorbance to protein concentration using

E""m

= 7.03 for pure inhibitor (4). The extinction coefficient,

EF^3m,

was de-rived from the measurement of absorbance of the solution of known weight of the purified inhibitor, which had been extensively dialyzed to remove salts, lyophilized, and then dried in a vacuum at 1 10°C over P205 for 4 days.

The concentrations of a2-M and

_a,-AT

in human plasma or serum were determined by the single radial immunodif-fusion technique using partigen plates (Behring-Werke AG).

Antigen-antibody-crossed electrophoresis was performed essentially according to Laurell (13). 30 ul of antigen solution was placed in a transverse slit (1.5 x 10 mm) in the agarose layer (0.8% agarose in barbital buffer, pH 8.6, ionic strength 0.05, and 1.8 mm thickness) on a glass slide (2.5 x 7.5 cm). The electrophoresis was performed for about 2Y2 h with a constant current of 2 mA/cm width. After-wards the gel slip was transferred to a second glass plate (5 x 7.5 cm). The remainder of the second plate was then covered with a layer of agarose containing 1% antiserum against a2-PI, 0.4% antiserum against plasminogen, or 2% antiserum against a,-AT, separately. The second electrophore-sis was performed by applying the electric current (1 mA/ cm width) at an angle of 900 to the direction of the first electrophoretic separation. The time of the second run was 10h. The buffer used for preparation of agarose gel and electrophoretic run was barbital buffer (pH 8.6, ionic strength 0.05). The plates were washed with saline to remove non-precipitated protein and dried. The immune precipitates were then stained with amido black.

'a2M'

antiplasmin activity. Antiplasmin activity of

plasma was estimated by the modified method of the fibrino-lytic assay described previously (1). 0.1 ml of 30

,ug/ml

(0.52 casein U/ml) plasmin in 25% glycerol was mixed and incubated at 37°C for 30 min with 0.1 ml of the test plasma which had been diluted 20-fold with barbital-buffered saline (5.2 mM barbital 0.14 M NaCl, pH 7.4). The mixture was then transferred to an ice-waterbath, diluted with 0.6 ml of cold barbital-buffered saline, and left for 3 min. Subse-quently, 0.1 ml of fibrinogen solution followed by 0.1 ml of

CONCE

2 4 6 8 IO 12

NTRATION OF INHIBITOR (mg/lOOml)

FIGUREI Areferencecurve for measuring theconcebtration

ofa2-PI by the one-dimensional method of Laurell (12). A linear relationship between concentration of purified

aL2-PI

and length ofprecipitate (peak heights) was obtained. For details see the text.

362

N.

Aoki,

M. Moroi,M.

Matsuda, and

K.

Tachiya

LU

c

t:

u LU cy-CL

LL

20 U/mlbovine thrombin (Parke, Davis & Company, Detroit,

Mich.) was added, and the solution was transferred to a 370C water bath, and the clot lysis time was recorded.

Fibrinogensolution used was 2% Cohn's fraction I of human plasma (Green Cross Corp., Osaka) which had been treated with lysine-Sepharose to remove plasminogen (14) and

contained1.6g/100 ml of fibrinogen. The control was run by replacing the test plasma with buffered saline. The differ-encebetween the control measurement and the test run, ex-pressed in fibrinolytic plasmin units as defined previously (1), divided by the volume (0.1 ml) of the test

sample,

multiplied by the dilution factor 20 represents the anti-plasmin units per milliliter. The determinations of the activityofplasma samples from 12 healthy adults revealed that the mean ±2 SD of normal values are 83±42 U/ml. The antiplasmin activity measured by this method is referred to

as a2-M antiplasmin activity, since the method measures mainly the antiplasmin activity of a2-M and is insensitive

to the change of a2-PI activity (1-3). Insensitivity of the method to the change of a2-PI may be explained by the fact that the method measures nearly total capacity of antiplasmin activity of whole plasma (1), and a2-PI contributes only 2.5%

onaverage of the total antiplasmin activity (see Discussion). Assay of

_ag-PI

activity. Activity of a2-PI in plasma or serum was assayed by the method based on the inhibitory

activity of plasma or serum to activator-induced clot lysis (1, 2). Activator-induced clot lysis is principally inhibited bya2-PI but also bya2-M toasmall degree (2). When plas-minogen-free plasma prepared by lysine-Sepharose (4)was

chromatographedonSephadex G-200 (PharmaciaFine Chem-icals, Div. of Pharmacia, Inc.) as described previously (2),

and each fraction ofthe effluentwasanalyzedforinhibitory

activity to urokinase-induced clot lysis (2), the a2-M frac-tionwhich emerged inthe first protein peak accounted for approximately 10-20% of the total activity of plasma

de-pendingona2-Mconcentration ineachplasma sampleused. The inhibitory activity exerted by the a2-M fraction was

diminishedtoapproximately halfby freezing and thawingthe

fraction, whereas inhibitory activity ofa2-PI fraction which

emerged

later inthechromatography was notaffected bya

freezing-thawing procedure.Thiswasalsoshownbythestudy

usingpurified a2-M (15) and purified a2-PI. Inhibitoractivity

of8mg/mla2-Mwas 120U/ml whenassayed bythe present

method for a2-PI activity. The activity was diminished to

55 U/ml by freezing and thawingthe same sampleof a2-M. However, a freezing-thawing procedure did not affect

ap-preciablythe inhibitor activity ofpurified a2-PI.

Therefore,

freezing and thawing the test sample before the assay

de-creasestheactivityby approximately 5-10lo and makes the assay method more specific to a2-PI. The decrease of

in-hibitory activity of a2-Mtoproteolytic enzymes(trypsin and

chymotrypsin) by freezingandthawing

procedures

has been

reportedalsoby others (16). Sincethe presence of

plasmino-gen in the test sample influences the assay method and giveserroneouslylowervaluesofinhibitor activity,

plasmin-ogen in the test sample was removed before the assay

by

lysine-Sepharose. 1 ml ofthe test plasma or serum, once

frozen and thawed, was applied on a column of lysine-Sepharose (1 ml)

equilibrated

withbarbital-buffered saline. The plasma wasallowed togo intothecolumn slowly and

wasfollowed with barbital-bufferedsaline. The first5mlof

theeffluent,which containedmore than 95% of theprotein

applied,wascollected. This fivefold-dilutedsamplewasused for the assay. An aliquot of the sample was diluted in the

test tube (10x75 mm) to 0.7 ml with cold

barbital-buffered saline and was mixed with 0.1 ml of2% Cohn's fraction Iofhumanplasma which contained 1.6g/100mlof

fibrinogenand 1.2 casein U/mlofplasminogen. Themixture

was cooled in an ice-water bath. To this mixture were added 0.1 ml of 600 CTA U/ml urokinase (Mochida Pharmaceutical Co., Tokyo), and 0.1 ml of 20 U/ml bovine thrombin to make a final clot of 1 ml. The tube was quickly shaken to mix the reagents well and placed in a 37°C water bath. A stopwatch was started at the time of addition of thrombin. Soon there formed a clot in which many small air bubbles werefonned and trapped. The clot lysis was made noticeable by the sudden rise of the air bubbles to the upper surfaces. In another set of experiments, similar clot lysis tests without test sample were performed with various concentrations of urokinase. A standard curve was constructed by plotting on double-logarithmic paper the clot lysis time obtained by various concentrations of urokinase against units of urokinase in a 1-ml clot. A straight line resulted over the range of 6-60 U urokinase/ml clot. Using this standard curve, clot lysis time obtained with a sample containing the inhibitor wasconverted to a value of urokinase activity. There was an inverse relationship between the amount of the inhibitor pres-ent and resulting residual urokinase activity, which is linear in the region from 60 urokinase U (control with urokinase and no inhibitor) to 20 U/ml clot (urokinase plus various amounts of inhibitor). The inhibitor units per milliliter plasma or serum were calculated as the difference between the measurement of urokinase activity in the presence and in the absence of the test sample, expressed as units per milliliter clot divided by the volume (milliliters) of the test sample introduced into the assay system and multiplied by 5 (the dilution factor derived from the initial treatment of plasma or serum with lysine-Sepharose). The mean ±SD of the activities of a2-PI in plasma from 29 healthy Japanese adults measured by this method were 1,224+374 U/ml.

Specificity

of the clot lysis method for

a2-PI.

The IgG fraction of rabbit antiserum to a2-PI or a2-M was added to lysine- Sepharose-treated and frozen-thawed plasma, in-cubated at 37°C for 1 h and further at 4°C for 24 h. The resulting precipitates were removed by centrifugation. The residual inhibitor activity in the supernate was measured by the clot lysis method. 2, 1, 0.5, and 0.25 mg of IgG fraction ofanti-a2-PI antiserum neutralized 95, 90, 82, and 70% of the inhibitor activity of 1 ml plasma, respectively. The similar amounts of IgG fraction of anti-a2-M antiserum did not ap-preciably affect the inhibitor activity. This indicates that the assay method is specific to some extent for a2-PI, but the ef-fect of any other proteinase inhibitor in plasma on the assay can not be excluded. Thrombin preparation used in the assay was crude material and might have been loaded with non-thrombin biologic activities. However, when purified thrombin prepared by the method of Lundblad (17) was used instead of crude thrombin, it gave the same results as thoseobtained with crude thrombin, suggesting that any con-taminant in the crude thrombin preparation did not affect the assay.

Radioiodination of plasminogen. Plasminogen was radio-iodinated by modification of the method of Greenwood et al. (18). To 13 mg of plasminogen in 6 ml of Tris-saline (0.05 M Tris-HCl, pH 7.4, 0.15 M NaCl) were added 0.15 mg of Nal in 0.15 ml of 0.05 M phosphate buffer (pH 7.4) and 80 ,uCi of 125I-Na solution (The Radiochemical Centre, Amersham, England) while stirring in an ice water bath. Immediately thereafter, 0.4 mg of chloramine-T (Tokyo Kasei Co., Ltd., Tokyo) in 0.1 ml of phosphate buffer was added, and the mixture was stirred for 1 min. Then 0.1 ml of Na2S202 solution (4 mg/ml phosphate buffer) was added to stop the reaction. Separation of

1251-plasminogen

from the reaction mixture was carried out by gel filtration through a column ofSephadex G-50. Elution was initially carried outwith 0.6 ml ofthe mixture of NaI (0.4 mg) and KI (2 mg), followed by

Tris-saline. Specific activity ofplasminogenbefore and after radioiodination were 23.1 and 20.2 U/mg protein, re-spectively.

Distribution of

'251-plasminogen

orplasminin

urokinase-activated plasma. 20 ul of radioiodinated plasminogen (0.91 mg/ml) was added to 2 ml plasma,

yielding

plasma containing9,ugradioiodinated plasminogen/ml.This plasma was activated with various amounts of urokinase and was

subjected to electrophoresis in agarose in the same way as

the firstelectrophoresisof

antigen-antibody-crossed

electro-phoresis. Immediately after the completion of electro-phoresis, the gel was cut transversely into 3-mm slices. Each slice was put into a counting vial, and radioactivity

was counted by Auto Well GammaSystem, Aloka JDC-752 (Aloka Co., Tokyo).

Statistical analysis. Statistical analysis was performed by Student'stanalysisforpairedorunpaired sampleswhere applicable. Forunpaired samples, Snedecor and Cochran's modification was applied (19). A P value>0.05 was

con-sidered to representastatisticallynonsignificantchange.The two-tailedtest wasused unless otherwise indicated.

Patients. Effects of intravenous infusion of urokinase on

plasmaplasmininhibitorswerestudiedon sevenpatients who received urokinase therapy within 72h after an attack of cerebral infarction. 240,000 U of urokinasewasadministered to each patient except foronepatientwho received 120,000U.

Patients having disseminated intravascular coagulation (DIC) with fibrinolysis were studied for the concentrations ofplasmin inhibitors. The diagnosis of DIC was based on

clinical observations as well as laboratory findings such as

increasedfibrin/fibrinogendegradation products

(16-640,ug/

ml, normal< 4,ug/ml [20]), decreasedplasma antithrombin III activity (10-80% of normal standard, normal range 90-110% [21]),prolonged prothrombin time (> 15

s,.normal

12-13 s [22]), decreased plasminogen (0.2-1.4 casein U/ml, normal range 1.4-2.0 casein U/ml), decreased fibrinogen

(20-200 mg/100 ml,normalrange 150-400mg/100 ml),and

decreasedplatelet count (<

150,000/,A,

normalrange

150,000-400,000/,ul)

measured by Coulter counter (Coulter Elec-tronics Inc., Hialeah, Fla.). These values were obtained by the repeated assays duringthe course of the diseases rather thanbyadeterminationon a single occasion. The gradual or

sudden decline of platelet count, fibrinogen, plasminogen, and antithrombin IIIactivity, togetherwithfurther

prolonga-tionof prothrombintimeand persistent presenceor asteady increase of fibrin/fibrinogen degradation products, was most suggestive of DIC.Underlyingdiseases were carcinoma,

acuteleukemia,and septicemia.

RESULTS

Development of the complexes of plasmin and

in-hibitors in

urokinase-activated

plasma in vitro.

Plasma

was

incubated

with various amounts of

uro-kinase, and subsequently subjected

to

antigen-anti-body-crossed electrophoresis.

Control

plasma without

addition of urokinase contained

a2-PI as a single

component

(Fig. 2A). When plasma was activated with

urokinase,

there appeared a new

component

which

possessed

the

antigenicity

of

a2-PI with a slower

mobility

in

the electrophoretic field

as

indicated

by arrow 1 (Fig. 2A). The new

component became

more

prominent,

and the a2-PI

component

itself was

dimin-ished

when the amount of urokinase was

increased,

as seen in the

bottom

figure of Fig. 2A. The new

component

disappeared

after the

plasma

wasabsorbed with

antiplasminogen

antiserum,

indicating

that it is

a

complex

between

a2-PI

and

plasmin

generated

by

urokinase. When

antigen-antibody-crossed

electro-phoresis

was carried out

using

antiplasminogen

anti-serum, the appearance of another new

_component

with a faster

mobility

was found in the

plasma

activated with

high

concentrations of urokinase

(>

100

U/ml)

as

indicated

_by

arrow 2 in

Fig.

2B. This

component became less

prominent

when the

_plasma

was absorbed

before

electrophoresis

with

anti-a2-M

antiserum,

indicating

that the component is the

com-plex

of

a2-M

and

plasmin.

It is

noteworthy

that the

complex

of

a2-M

and

plasmin

did not

develop

when the concentration of

urokinase

was as low as 25 U/ml

(Fig. 2B),

whereas the formation of the

complex

of

a2-PI

and

plasmin

was observed

(Fig.

2A).

FIGURE 2 _{Formation of the complex of the plasma} inhibi-tors and plasmin in plasma activated with urokinase in

vitro, demonstrated by antigen-antibody-crossed

electro-phoresis. To each 1 ml of plasma was added 2.5 and 50

pA of urokinase solution (10,000 U/ml) separately, and the mixtures were incubatedat37°Cfor 2 hand further atroom

temperature foranadditional 12 h. They were subsequently

subjectedto_{antigen-antibody crossed electrophoresis using} antiseraagainst a2-PI (A) orplasminogen (B). The direction ofthe first electrophoresis was from the left to the right.The amounts (units)ofurokinase added to 1 ml plasma are

indi-cated between the panels. Arrows indicate the complex of

plasmin and inhibitors. Arrow 1, the complex of a2-PI and plasmin. Arrow 2, the complex of a2-M and plasmin.

364

N. Aoki, M. Moroi,

M.

Matsuda, and

K.

Tachiya

RR',

gmmm

:.un

When these urokinase-activated plasma samples were

analyzed by

the crossed

electrophoresis

using antiserum against

_a,-AT,

no

formation

of the complex

of

a1-AT

and plasmin

was

observed,

even in plasma

activated with high

concentrations of urokinase, such as 500

U/ml plasma.

In these plasma samples, nearly all plasminogen was converted to plasmin, and most of the a2-PI

formed

complexes with plasmin (Fig. 2). The formation of the complex of

a,-AT

and plasmin was observed only after the addition of exogenous

plasmin

tothis highly

activated

plasma.

Plasma containing radioiodinated plasminogen was

activated

with

urokinase, and

the

distribution of

radioactivity in plasma fractions was analyzed by

electrophoresis.

The

results

are shown in Fig. 3. When

plasma

was

activated

with 25 U of urokinase/ml plasma, the

peak

of

radioactivity shifted

to the location (slice 5)

corresponding

to

the peak of a2-PI-plasmin

complex indicated by

arrow 1 in Fig. 2A (Fig. 3A). The

peak of a2-PI-plasmin complex became

prominent with concurrent decrease of the original plasminogen

peak (slice

3) when

amounts

of urokinase used

were

increased

(Fig. 3B). With increasing

urokinase,

there

developed another

new

peak of radioactivity (slice

9) which

corresponds

to the location of the

peak

indi-300

X200

A t

u 100 r

0

30

p

In

B

10

CONTROL UK 25UNITS

2 4 6 8 10 12 14 SLICE NUMBERS

0o-K125UNITS

nI

.

1

U 2 4 6 8 10 12 14

SLICE NUMBERS

FIGuRE 3 Distribution ofradioactivity in plasma fractions

after activation of plasma containing radioiodinated

plas-minogen with urokinase.Toeach 1mlofplasmawasadded

2.5, 12.5, and 50 ,ul of urokinase solution (10,000 U/ml)

separately, andthemixtureswere incubated at 37°C for2h and further at room temperature for additional 12h. They

were subsequently subjected to electrophoresis in agarose,

and the distribution of radioactivity was analyzed. The amounts ofurokinase addedto 1 mlplasmaareindicatedon each figure. Control was run withoutaddition of urokinase. Slices of agarose plate are numbered from the origins wherethe samples wereapplied. Details of the methodare describedin thetext. UK, urokinase.

>o[

-NoNDEPLETION

< IOO X2M_DEPLETED

2 4 6 8 10 12 14

SLICE NUMBERS

FIGURE 4 Depletion of plasmin-a2-M complex by

anti-serumagainsta2-M. 0.1 ml ofantiserum was addedto 1 ml

plasmawhichcontained radioiodinatedplasminogenandhad been incubated with 125 U/ml urokinase as in Fig. 3. The mixture was then incubated at 37°C for 30 min and further at40C for additional 12h. The resulting precipitates were

removed by centrifugation andthe supernatewas subjected

toelectrophoresis. Thecondition of the electrophoresis and the subsequent treatment were the same as in Fig. 3. The

control(nondepletion)was runby replacing rabbitantiserum

with unimmunized rabbit serum.

cated by

arrow 2 in Fig. 2B (Fig. 3B). This new peak is most likely

attributable

to the formation of the a2-M-plasmin conmplex, in that the peak was destroyed by the treatment of the plasma with a2-M anti-serum before electrophoresis (Fig. 4).

Development of

the

complexes

of

plasmin

and

inhibitors

in

plasma

of patients receiving urokinase intravenously. For therapeutic purposes 240,000 or 120,000 U

of

urokinase together with 5,000 U of heparin was

infused intravenously for

2h to seven

different

patients with

cerebral

infarction.

Immedi-ately before the infusion, and

at 0, 3, 6

and

18h

after the

termination

of the

infusion, blood samples

were

withdrawn

from the patients

and subjected

to

analyses by

antigen-antibody-crossed electrophoresis (Fig. 5). The

complex of

a2-PI

and plasmin

was

ob-served

prominently

inthe

samples withdrawn

immedi-ately

after the infusion (Fig. 5). The

complex

became less prominent at3h, andno

complex

was found at6h after the infuision (Fig. 5). The antigen concentration,

as well as the activity of a2-PI,

decreased

concur-rently

with the appearance of the

complex of

a2-PI

and

plasmin.

The decrease of activity was more

re-markable than

that

of

antigenconcentration at the time

immediately after

the infusion (Fig. 5). The

complex

of

plasmin and

a2-M or

a,-AT

as

described

in in

vitro

studies

mentioned above

was not

found

at any time.

Decrease

of

a2-PI after the

intravenous

infusion

of

urokinase. To continue our

studies

in these patients,

concentrations (antigen) of the inhibitors,

a2-PI,

a2-M, and

a,-AT,

in plasma were measured

immuno-chemically

before and after the intravenous infusion (2 h) of 240,000 U of urokinase together with 5,000 U

of heparin. Inhibitory activity of

a2-PI,

a2-M

anti-plasmin activity, anti-plasminogen, and fibrinogen were also

assayed.

The concentrations and the activity of

E _a

(a.

-11.

FIGuRE 5 Formationand disappearance of the complex of a2-PI and plasminin plasma ofthe patientsunder urokinasetreatment,demonstratedbyantigen-antibody-crossed electrophoresis

using antiserum againsta2-PI. Plasma sampleswerewithdrawn from thepatientsbefore (a),

im-mediately after (b), and6hafter (c) the infusion of urokinase.Patients AandBreceived 120,000 and 240,000 U ofurokinase, respectively. Arrows indicate the complex ofa2-PI and plasmin. The concentration (mgantigen/100 ml) and the activity (U/ml shown in parentheses) ofa2-PI are indicated under each figure.

a2-PI were

significantly decreased by the infusion of

urokinase

and had not yet returned

to

the

original

values when

examined at 18 h after the termination of

infusion (Tables

I

and II). On the

contrary, the

con-centrations

of

a2-M and a1-AT were not changed

signif-icantly by the infusion. Some significant increase of

a2-M

antiplasmin activity was noticed at the

time

im-mediately after the infusion (Table II). Plasminogen

TABLE I

ChangesintheConcentrations ofInhibitors inPlasma aftera 2-hUrokinaseInfusion

Timeafterthe start ofinfusion, h

Inhibitors 0 2 20

mg/100ml

a2-PI 4.78±+1.27* 3.42+0.52 3.02±0.29

P<0.02t P<0.02

a2-M 191+22 191+19 198+26

NS NS

a,-AT

244+40 248+41 261+24

NS NS

* Mean+SD

(n.=

6).

t P valuesreferto the difference in concentrations insamples

at2or 20 hfrom those in samplesat 0 time.

decreased significantly after urokinase infusion,

whereas fibrinogen remained unchanged (Table III).

Decrease of

atrPI

in DIC with fibrinolysis.

Con-centrations

of

a2-PI, a2-M,

and

a,-AT

in

plasma

were

measured immunochemically in patients having DIC with

fibrinolysis, and

compared

with

normal

values.

The

mean +2 SD

of

the concentrations

of

a2-PI in

plasma

from 25

healthy

Japanese

adults

were 5.98 ±+1.76 mg/100 ml. The

reported values of

the mean

±2

SD of concentrations

of

a2-M and

a,-AT

in

plasma

TABLE II

ChangesinInhibitorActivitiesofPlasma after

a 2-h Urokinase Infusion

Timeafter thestartof infusion, h

Inhibitors 0 2 20

U/mi

a2-PI 1,461+298* 588±206 911+244

P<0.001t P<0.005

a2-M 103+22

136±

18 95+22

antiplasmins P < 0.05 NS

* Mean+SD

(a2-PI,

n =6; a2-M, n=5).

t P values refer to the difference in activities in samples at 2or 20 h from those in samples at 0 time.

366

N.Aoki, M. Moroi, M.

Matsuda,

andK. Tachiya

TABLE III

Changes intheConcentrationsof Plasminogen and FibrinogeninPlasma aftera2-h

Urokinase Infusion

Time afterthestart ofinfusion, h

0 2 20

Plasminogen,

casein U/ml 1.41±0.26* 1.06±0.22 1.05±0.3

P<0.0005t P<0.05

Fibrinogen,

mg/lOO

ml 394±123 337±91 346±86

NS NS

*Mean±SD(n =5).

t Pvalues refertothe differencein concentrations in

samples

at2or20 hfrom thosein

samples

at0time.One-tailedtest.

from

160

healthy Japanese

adults measured by single

radial

immunodiffusion are

210+100

and

200+100

mg/100

ml, respectively (23). The concentrations of

a2-PI

in most

of the patients

were

found

to

be lower

than the

normal range, whereas values of a2-M

were in

the

normal range except for

one

case, and those

of

a,-AT

were

higher

in

many

cases

than the normal

range (Fig. 6). Many of the values of a2-M,

even

though

in

the

normal range,

were

well below the

mean

level of normals, thus

implying that there had been

some

reduction

in

the

level

in

those

cases. No

com-plex of

plasmin

with any

one

of the inhibitors

was

detected when these

patients' plasmas

were

analyzed

by

antigen-antibody-crossed

electrophoresis.

a2-PI

0

C!4

E

6

2

Ot2-M

800

600 300

-0

*0 0

-40-0

*t

O1

200-

1001-OL

P<0.001

0

*.0

I

---a___

4001

2001-o0 NS

FIGURE 6 Concentrations of a2-PI, a2-M, and al-AT in

plasma ofthe patients with DIC, measuredbythe

immuno-chemical method. Themean concentrationofeach inhibitor is indicatedbyashorthorizontalline. Themeans +2 SDof

normal values forhealthyadults are indicated bysolidand

interrupted horizontallines,respectively.Pvalues refertothe difference in concentrations in DIC from those in normal.

DISCUSSION

The concentration of a2-PI in normal plasma was

found

to

be approximately 5-7 mg/100 ml, which is calculated

to

be _0.7-1.0 ,uM on the basis of mol wt 67,000.

The molar concentrations of a2-M (mol wt 820,000) and

al-AT (mol wt 54,000) in normal

plasma are estimated

on the

average at 2.4 and 37

,uM,

respectively, while

their average concentrations are 210 and 200 mg/100

ml, respectively (23). Consequently, it may be

con-cluded

on

the assumption

of one-to-one molar reaction

of

inhibitor and enzyme that a2-PI contributes only

2.5% on average of the

total capacity of antiplasmin

activity of the

whole plasma. However, a2-PI has the

strongest

affinity for plasmin among these inhibitors.

It had been suggested that the strong affinity of a2-PI

for

plasmin plays an important role in the regulation

of

fibrinolysis. To test this possibility, the behavior of

a2-PI

in

fibrinolytic states was explored in the present

study.

When plasminogen in

plasma was activated in vitro

by

a

relatively low concentration of urokinase, there

appeared readily

a

plasmin-a2-PI complex. However,

no

complex of plasmin-a2-M or

a,-AT

was

observed.

The formation of the plasmin-a2-M complex was

ob-served

only when plasma

was

activated with a larger

amount

of urokinase and after

most

of the a2-PI

was

consumed by forming complexes with plasmin.

The formation of

plasmin-al-AT

complex was not

ob-served

even

when

all of the plasminogen in plasma

was

converted to plasmin unless exogenous plasmin

was

added.

When a relatively low

dose of urokinase was infused

intravenously

to

patients, the consistent

findings were

the

decrease of the concentration and the activity of

a2-PI

and the

concomitant

formation of the

plasmin-a2-PI

complex. The decrease of the activity of

a2-PI

was more

remarkable than that of the

con-centration at

the

time

of the

existence

of the

complex

in

blood. This

can

be

explained by the fact that the

concentration

measured

by the immunochemical

method includes the enzyme-inhibitor complex which

has

lost the activity.

In contrast

to a2-PI,

concentrations

of a2-M and

a,-AT

were not

significantly changed, and

no

complex

of plasmin-a2-M

or

-a,-AT

was

observed.

A

significant increase of a2-M antiplasmin activity

observed

immediately

after the

infusion

might

be

at-tributable

to

antiplasmin

activity of

heparin-anti-thrombin

III

(24), because there

was no

significant

increase

of a2-M antigen

concentration

throughout

the

study, and

no

significant

increase

of

a2-M

antiplasmin

activity

was

observed

18h

after

the

termination

of

infusion

when no

residual effect of heparin

was

de-tected.

Arnesen and

Fagerhol

studied effects of urokinase

fusion

ona2-M,

a,-AT,

andantithrombin III inplasma (25). The doses of urokinase that

they

used were far greater than

those

used

in the present

study.

They infused

nearly twofold

the amount of

urokinase

as

compared

to ours in 10 min as a

loading

dose and continued infusion of a maintenance dose for

18h.

Their

hourly

maintenance dose is

nearly

equivalent

to

the total

amount

of

urokinase infused

into each patient in

the

present

study.

In spite

of their large

dose, the decrease of

a2-M after

completion

of

in-fusion

was

only

20% on average of the

preinfusion

values.

There

was no

change of

plasma

antithrombin III

level, and the

concentration

of a,-AT

was even

increased. Nilehn and

Ganrot

activated practically

all

plasminogen by infusing

a

large dose of

strepto-kinase

and observed the decrease of

a2-M to 50% of

the initial

values

(26). In the present

study,

the de-crease

of

plasminogen level averaged

25% of

the

pre-infusion values, and

no

decrease

of a2-M was

ob-served. These data, together

with

those

obtained by

others (27, 28), indicate that a2-M

is one

of the

major

antiplasmins which

reactwith

plasmin

generated

dur-ing

fibrinolytic therapy.

The present

study provides,

furthermore, information

on a

newly discovered

in-hibitor

in

addition

to

a2-M.

The

present

study

suggests that a2-PI is the most

reactive

plasmin

inhibitor

in

plasma, and whenever

plasminogen

activation

takes

place,

a2-PI is

the

first

inhibitor

which reacts with

plasmin formed. Only

when

the

excess amount

of

plasmin

is

formed, and when

most a2-PI

has been consumed by forming complexes

with

plasmin,

then a2-M will

become

the

principal

inhibitor

to

plasmin. This

view is

supported by the

finding that,

among

the

major

plasmin inhibitors

in

plasma,

a2-PI was

the

only inhibitor of

which

the

concentration was

significantly decreased

in patients with DIC. However, no

complex

of

plasmin-a2-PI

was

observed

in

plasma of these

patients, at

least

by

the

method presently employed.

This is

explained by

the

rapid

disappearance

of

plasmin-a2-PI complex from

the

circulating blood, which,

in the present

study,

was

indeed shown after the infusion of urokinase.

The

plasmin-a2-PI complex formed after intravenous

in-fusion of 240,000

U

of urokinase

disappeared

from the

circulating blood

within 6h. The

complex

was

proba-bly removed by the

clearance

mechanism

of the

retic-uloendothelial

systeminorgans

especially

inthe

liver,

since the

complex

is

formed by

a

stable covalent

bond (4), and its

dissociation

can not

readily

occur.

a2-M might

also

have reacted with plasmin in these DIC patients, inasmuch as almost all values of a2-M were

below

the mean value

although

they were in the normal range. However, this possibility was not

sta-tistically

proved in the present study (P> 0.1). The cause ofthe increase of

a,-AT

observed

in these DIC

patients

was not

known,

but it is

interesting

to note

that

_a,-AT

concentration in

plasma

was

steadily

increasing

during prolonged

infusion of urokinase

(25).

Because the

molecular weight of

the complex of

a2-PI

and the

light

chain of

plasmin

was

previously

reported

as

84,000

(4)

and

the molecular weight

of the

heavy

chain was

nearly 67,000 (4),

the molecular

weight

of

plasmin-a2-PI complex

must be close to

150,000.

The

plasmin-a2-PI complex might be

the

same as the

complex reported by

Collen et al.

(29).

They reported that, during

the activation of

plasmin-ogenin

plasma

with urokinaseinvitro, there

appeared

the

complex

of

plasmin

and a

hitherto

unidentified

antiplasmin.

The molecular

weight

of this

complex

was

estimated

by gel

filtration to be in the region of 130,000-170,000, which is the same region of

the

estimate of molecular

weight

of

plasmin-a2-PI

com-plex.

Miillertz

(30)

also reported that

urokinase-activated

plasma contained,

ata low concentration of

urokinase,

a component with

plasmin

+

plasminogen

antigenicity which was eluted with

nearly

7S protein from

Sephadex

G-200 and had

8-mobility by

gel

electrophoresis.

This component appears from its

phys-ical properties to be identphys-ical with

plasmin-a2-PI

complex.

Inaddition to these

observations,

plasma inhibitors of

fibrinolysis

which possess properties similar to those of a2-PI were recently described (31-35). The question

as to whetherornota2-PI is identical with any one

of

these inhibitors needs further

investigation, including

elucidation of the

physicochemical

and

immunochem-ical properties of each inhibitor.

Crossed

immunoelectrophoresis

of some

plasma

samples

using anti-a2-PI antisera gave double immuno-precipitate lines as seen in patient B in Fig. 5. Development of

double immunoprecipitate lines

might be due to the molecular

heterogeneity

of antigen as suggested by Laurell (36), and the antisera

might

contain

antibodies

against more than one of the determinants. Molecular

heterogeneities

of a2-M and

al-AT

have

already

been revealed by crossed

im-munoelectrophoresis

(37). Another possibility that the antiserum is not specific and reacts with two different

molecules

which possess the same electrophoretic

mobility

cannotbe

excluded,

although

immunoelectro-phoresis

and

double immunodiffusion

have not been able to

demonstrate

this possibility (4).

ACKNOWLE DGMENTS

Theauthors would like to thank Dr. M. Yoshida and Dr. A.

Ueki of the Department of Neurology at the Jichi Medical

School fortheir generous cooperation. Acknowledgment is

also due to Dr. W. H. Seegers, Wayne State University, for

hiscritical reading.

This work was supported bya grantfrom the Ministry of Education of the Govemment of Japan.

REFERENCES

1. Aoki, N., and K. N. von Kaulla. 1971. Human serum plasminogen antiactivator: its distinction from anti-plasmin. Am. J. Physiol. 220: 1137-1145.

2. Aoki, N., and T. Kawano. 1972. Inhibition of plasminogen activators by naturally occurring inhibitorsinman. Am.J.

Physiol. 223: 1334-1337.

3. Aoki, N., and M. Moroi. 1974. Distinction of serum

in-hibitor of activator-induced clot lysis from

a,-anti-trypsin. Proc. Soc. Exp. Btol. Med. 146: 567-570. 4. Moroi, M.,and N. Aoki. 1976. Isolation and

characteriza-tionof a2-plasmin inhibitor from human plasma. A novel proteinaseinhibitor which inhibits activator-induced clot lysis. J. Biol. Chem. 251: 5956-5965.

5. Brockway, W. J., and F. J. Castellino. 1972. Measure-ment of the binding of antifibrinolytic amino acids to variousplasminogens. Arch. Biochem. Biophys. 151: 194-199.

6. Robbins, K. C., and L. Summaria. 1970. Human plasmin-ogen andplasmin. Methods Enzymol. 19: 184-199. 7. Barlow, G. H., L. Summaria, and K. C. Robbins. 1969.

Molecular weight studies on human plasminogen and plasmin at the microgram level. J. Biol. Chem. 244: 1138-1141.

8. Alkjaersig, N.,A. P. Fletcher, and S. Sherry. 1959. The mechanism of clot dissolution by plasmin.

J.

Clin. Invest. 38: 1086-1095.

9. Remmert, L. F., and P. P. Cohen. 1949. Partial puri-ficationand properties ofaproteolytic enzyme of human serum.J. Biol. Chem. 181: 431-448.

10. Ratnoff, 0. D., and C. Menzie. 1951. A new method for the determination of fibrinogen in small samples of plasma. J. Lab. Clin. Med.37: 316-320.

11. Rowe, D. S., and J. L. Fahey. 1965. A new class of human immunoglobulins. II. Normal serum IgD. J. Exp. Med.

121: 185-199.

12. Laurell, C. B. 1966. Quantitative estimation of proteins

by electrophoresis in agarosegel containing antibodies. Anal. Biochem. 15: 45-52.

13. Laurell, C. B. 1965. Antigen-antibody crossed electro-phoresis. Anal.Biochem. 10: 358-361.

14. Matsuda, M., S. Iwanaga, and S. Nakamura. 1972. A

simple, large scale method for preparation of

plasmino-gen-free fibrinogen.Thromb. Res. 1: 619-630.

15. Iwamoto, M.,andY.Abiko. 1970. Plasminogen-plasmin

system. Preparation of a2-macroglobulin antiplasmin from human plasma. Biochim. Biophys. Acta. 214: 402-410.

16. Lebreton de Vonne, T., H. Mouray, G. Berthillier, and

R.Got. 1972.Influencedu vieillissement des a2-macro-globulines de lapin sur la formation du complexe a2-macroglobuline-enzymes. Biochim. Biophys. Acta. 257:

365-371.

17. Lundblad, R. L. 1971. A rapid method for the

purifica-tion of bovine thrombin and the inhibition of the puri-fied enzyme with phenylmethylsulfonyl fluoride.

Bio-chemistry. 10:2501-2506.

18. Greenwood, F. C., W. M. Hunter, and J. S. Glover. 1963. The preparation of 1311-labelled human growth hormone of high specific radioactivity. Biochem. J. 89:

114-123.

19. Snedecor, G. W., and W. G. Cochran. 1967. Statistical Methods. IowaState University Press, Ames, Iowa. 6th

edition. 114-116.

20. Merskey, C., G. J. Kleiner, and A. J. Johnson. 1966. Quantitative estimation of split products of fibrinogen

in human serum, relation to diagnosis and treatment.

Blood. 28: 1-18.

21. Monkhouse, F. C. 1963. The influence of inorganic salts

on plasma antithrombin activity. Thromb. Diath. Haemorrh.9: 387-394.

22. Tocantins, L. M. 1964. Estimation of prothrombin

(One-stage method of Quick).In Blood Coagulation,

Hemor-rhage and Thrombosis. L. M. Tocantins and L. A. Kazal, editors. Grune & Stratton, Inc., New York. 148-150.

23. Matsuda, T. 1967. Coagulation studies on strokes. Nihon Rinsho. 34: 72-78.

24. Highsmith,R.F., andR. D.Rosenberg.1974.The

inhibi-tion of humanplasmin by human antithrombin-heparin cofactor.J. Biol. Chem. 249:4335-4338.

25. Arnesen, H., and M. K. Fagerhol. 1972.c2-Macroglobulin, a1-antitrypsin, and antithrombin III in plasma and serum during fibrinolytic therapy with urokinase. Scand. J. Clin. Lab. Invest. 29:259-263.

26. Nilehn, J-E., and P. 0. Ganrot. 1967. Plasmin, plasmin inhibitors anddegradation products of fibrinogenin

hu-man serum during and after intravenous infusion of streptokinase. Scand. J. Clin. Lab. Invest.20: 113-121.

27. Fischer, M. 1971. Influence of thrombolytic therapy on human antiplasmins. Thromb. Diath. Haemorrh. 47

(Suppl.): 153-157.

28. Spottl, F., and F. Holzknecht. 1970. The influence of

inhibitors of plasmin and plasminogen activation onthe

streptokinase-induced fibrinolytic state. Thromb. Diath. Haemorrh.24: 100-112.

29. Collen, D., F. De Cock, and M. Verstraete. 1975.

Immunochemical distinction between antiplasmin and

alpha1-antitrypsin.Thromb.Res. 7: 245-249.

30. Mullertz, S. 1974.Different molecular forms of plasmin-ogen and plasmin produced by urokinase in human

plasma and their relation to protease inhibitors and lysis of fibrinogen and fibrin. Biochem. J. 143: 273-283. 31. Hedner,U. 1973.Studieson aninhibitor of plasminogen

activation in human serum. Thromb. Diath. Haemorrh. 30: 414-424.

32. Mui,P. T.K.,H. L.James, and P.Ganguly. 1975. Isola-tion and properties of a low molecular weight

anti-plasmin of human blood platelets and serum. Br. J.

Haematol.29: 627-637.

33. Gallimore, M. J. 1975. Serum inhibitors offibrinolysis.

Br.J. Haematol.31: 217-231.

34. Hedner, U., and D. Collen. 1976. Immunochemical

dis-tinction between the inhibitors of plasminogen activa-tion and antiplasmin in human plasma. Thromb. Res. 8: 875-879.

35. Edy, J., F. De Cock, and D. Collen. 1976. Inhibition ofplasmin by normal and antiplasmin-depleted human plasma. Thromb. Res. 8: 513-518.

36. Laurell, C. B. 1972.

Electroimmunoassay.

Scand._J.Clin.

Lab.Invest. 29(Suppl. 124):21-37.

37. Ganrot, P. 0. 1972. Crossed immunoelectrophoresis.

Scand.J.Clin.Lab. Invest. 29(Suppl. 124): 39-47.